@f2harrell
@learningonco @KamravaMD @5_utr Forest plots based on subgrouping should be banned. Should always be model-based, treat continuous variables as continuous, and discount interaction effects using Bayesian shrinkage priors.
